wOBA is really flawed stat

[iAugust]

Take Stealing out (that's not a hitting stat, it's a running stat)

Ask and you shall receive.

http://www.fangraphs.com/blogs/index...nning-changes/

[bosox3431]

Yeah, the problem I have with the formula is that by defenition wOBA is supposed to tell you how many runs you produce per at bat. The thing is I view HBP and stealing stats a non batter's statistic, and HBP especialy a non offensive statistic (that's a pitchers error in my view).

If they said wOBA was offence produced by a player's hitting skills and on base running skills then it would be more acceptable for me. But it's clearly saying runs produced at bat by the hitter, not a player's hitting and base running skills.

Steaking still helps create runs if the batter is successful enough. Its a skill. I can understand hbp though.

[iAugust]

Stealing still helps create runs if the batter is successful enough. Its a skill. I can understand hbp though.

I agree with this which is why I'm a little upset that FanGraphs removed it from wRC. I understand taking it out of wOBA to make wOBA a purely batting metric, but we have the weighted run value of a steal so why remove it from wRC too? Not a big fan of the change

[jej]

I'm still not completely sure what he thinks is flawed. He keeps bringing random incorrect information up.

SB were taken out. The difference between HBP and uBB is not large enough to me to call it a completely flawed stat, especially when it's competition are 2 counting stats useless average that tells me one thing.

They get the numbers from other data. They don't just pull it out. HBP must be more valuable in their research.

[flips333]

It doesn't matter is the stat is 2.456374 x the density of the hitter's poop. What matters is if it predicts runs and predicts wins significantly better than simpler stats.

[Milnertime]

It doesn't matter is the stat is 2.456374 x the density of the hitter's poop. What matters is if it predicts runs and predicts wins significantly better than simpler stats.

Exactly this.

Correlation to real runs is the entire point. People seem to forget that, quite often.